COUNCIL OF THE EUROPEAN UNION. Brussels, 21 June 2001 (26.09) (OR. fr) 9269/01 ASILE 31

COUNCIL OF THE EUROPEAN UNION Brussels, 21 June 2001 (26.09) (OR. fr) 9269/01 ASILE 31 COVER NOTE from : Mr Bernhard ZEPTER, Deputy Secretary-General of the European Commission date of receipt : 14 June 2001 to : Mr Javier SOLANA, Secretary-General/High Representative Subject : Commission staff working paper: "Evaluation of the Dublin Convention" Delegations will find attached Commission document SEC(2001) 756. Encl.: SEC(2001) 756 9269/01 mip/mck/jr 1 DG H I EN

COMMISSION OF THE EUROPEAN COMMUNITIES Brussels, 12.06.2001 SEC(2001) 756 COMMISSION STAFF WORKING PAPER EVALUATION OF THE DUBLIN CONVENTION

This document evaluates the application of the Dublin Convention. The idea of such an evaluation was adopted in the Vienna Action Plan, i and the Commission agreed to carry it out. INTRODUCTION...1 I. STATISTICAL OUTLINE...2 II. CRITERIA FOR ALLOCATING RESPONSIBILITY AND THE PRODUCTION OF EVIDENCE...4 1. Establishment of initial responsibility...4 2. Taking back...7 3. Requests for information under Article 15...8 4. Exemption clauses...8 III. TIME LIMITS...9 1. Time limits laid down by the Convention and by decision No 1/97 of the Article 18 Committee of 9 September 1997...9 2. Other sources of time limits...10 3. Consequences of time limits...10 4. Role of bilateral agreements...11 IV. LEGAL QUESTIONS...12 1. Vagueness of, and gaps in, the Convention...12 a) Scope12 b) Definitions...13 c) Wording of the provisions...13 d) Gaps in the Convention...14 2. Action by national courts...14 V. PRACTICAL ARRANGEMENTS...16 1. Communications...16 2. Transfers...16 VI. ROLE OF THE CONVENTION AS A MEASURE COMPLEMENTING FREEDOM OF MOVEMENT AND ITS INFLUENCE ON THE DISTRIBUTION OF ASYLUM SEEKERS IN THE EUROPEAN UNION...17 VII. ADMINISTRATIVE BURDEN AND COSTS...18 1. Workload...19 2. Costs...19 Endnotes...21 I

INTRODUCTION The Convention determining the State responsible for examining for asylum lodged in one of the Member States in the European Communities was signed in Dublin on 15 June 1990 and came into force on 1 September 1997. The Convention is designed to allocate, to a single Member State and on the basis of objective criteria, the responsibility for examining an asylum application submitted in a Member State and thus to ensure that the of all asylum seekers will be examined and that they will not be left in prolonged uncertainty as to whether their application will be accepted or not. The general principle underlying the criteria for allocating responsibility is that, in an area where the free movement of individuals is guaranteed under the Treaty establishing the European Community, each Member State is answerable to all the others for its actions or shortcomings as regards the entry and residence of third-country nationals. The criteria set by the Convention are therefore designed to allocate responsibility for examining an asylum application to that Member State which has played the most important part in the entry or residence of the person concerned. Under the Convention the responsible Member State has a number of obligations, including that of seeing through the examination of the asylum application and that of taking of the applicant if the latter goes illegally to another Member State. The Convention also lays down arrangements and deadlines for submitting the whereby Member States ask each other to take asylum seekers in or to take them back. A ministerial committee set up by Article 18 of the Convention is responsible for examining any questions relating to implementation or interpretation. This committee has adopted various decisions laying down recommendations and guidelines for improving the operation of the Convention; in particular it has shortened the deadlines and clarified certain concepts contained in the Convention. Article 63(1)(a) of the EC Treaty, as amended by the Treaty of Amsterdam, requires the Council to adopt an instrument of Community law laying down criteria and mechanisms for determining which Member State is responsible for considering an application for asylum submitted by a national of a third country in one of the Member States. This instrument will replace the Dublin Convention. Before drafting any new measures, however, the Council felt that the operation of the Convention should be evaluated, a task which the Commission agreed to take on. At the same time as this evaluation, the Commission initiated a wide-ranging debate on the future prospects for the instrument which would determine the Member State responsible. To this end, it distributed a working paper entitled "Revisiting the Dublin Convention: developing Community legislation for determining which Member State is responsible for considering an application for asylum submitted in one of the Member States" (SEC(2000)522), which has been scrutinised and discussed within the Council and on which numerous interested bodies have submitted their observations. 1

I. STATISTICAL OUTLINE An initial picture of the application of the Dublin Convention can be gained from statistical data. This provides a quantitative perspective from which some lessons can be learned. Given the variety of methods by which the Member States statistics are recorded and their uneven level of detail, the figures available indicate orders of magnitude only and do not claim to be scientifically precise. The data collected for this evaluation are for 1998 and 1999 (tables I and II). To smooth the aberrations found in the annual data (due, for example, to slippage from the first year to the second) and to obtain more reliable data from a wider statistical sample, it was thought preferable to base the evaluation, wherever possible, on the cumulative data for the two reference years (although this means that, in some cases, estimates have had to be made to compensate for missing data - see Table III). ii Consequently, such conclusions as it has been felt possible to draw in this report must be treated with caution the gaps in the data and the empirical method used are both sources of uncertainty. The crude data on the application of the Convention throughout the EU, aggregated over a period of two years, can be summarised as follows: Table 1 : Aggregate data EU 15, 1998-99 Application of the Dublin Convention over the 24-month period Total number of to take /take back submitted to other Member States Proportion of the total number of for asylum (II/I) (%) No. of to take /take back accepted Acceptances as a proportion of the number of to take /take back submitted (IV/II) (%) Acceptances as a proportion of the total number of for asylum (IV/I) (%) No. of asylum seekers actually transferred Actual transfers as a proportion of accepted to take /take back (VII/IV) (%) as a proportion of total to take /take back sent to other MS (VII/II) (%) Actual transfers as a proportion of the total number of for asylum (VII/I) I II III IV V VI VII VIII IX X EU : of for asylum: 1998-99 655 204 EU aggregate 39 521 6.00 27 588 69.80 4.20 10 998 39.90 27.80 1.70 One is immediately struck by the small proportion of asylum giving rise to a request to another Member State to take of/take back an applicant. While, in theory, every asylum application is examined, however briefly, in the light of the responsibility criteria in the Convention, only in 6% of cases does the Member State with which the asylum application is lodged request another Member State to take of/take back the applicant. The second immediate finding is the high success rate for addressed to other Member States under the Convention. Since the rate achieved is close to 70%, it may be concluded that, most of the time, are only submitted advisedly and that, generally, Member States examine the submitted to them by other Member States in good faith and in a positive frame of mind. However, given the relatively low number of submitted, the proportion of asylum where a Member State other than that in which the application was lodged is held to be responsible is only 4.20%. The result is that in more than 95% of cases it is the Member State in which the asylum application is lodged which assumes responsibility for examining it. (%) 2

The third finding concerns the gap between the number of cases where a transfer has been agreed and the number of transfers actually carried out or recorded. The transfer of the asylum seeker is attested only in slightly less than 40% of cases; the proportion of transferable asylum seekers not transferred is thus 60%. There are undoubtedly a number of voluntary transfers which take place without being recorded, and changes in status (waiver of the asylum application, grant of residence on other grounds, etc.), transfers postponed for legal or humanitarian reasons and other situations likely to obstruct the transfer have to be taken into account, so that the scale of the problem is probably less than the statistics show. Nevertheless, a certain evaporation occurs, resulting in the creation of a pool of aliens, most often in an unlawful situation, who have expressed their intention to apply for asylum but whose application will not be examined either in the Member State where they lodged it, which is not responsible, or in the Member State responsible, to which they have not travelled. The above summary table shows that the number of asylum seekers actually transferred accounts for slightly less than 30% of those in respect of whom a request to take /take back is made, or for 1.70% of the total for asylum lodged in the 15 Member States. These overall figures should not hide the variation in the results for the individual Member States (see Tables I, II and III). It will be seen straightaway that the Member States fall into two groups, depending on whether the balance of the transfers they make and those they receive is positive or negative. Although the data on inward transfers is incomplete, Germany, Austria, Italy and France are clearly the States where the balance, in absolute terms, is the most unfavourable, followed by Spain, Greece, Portugal and Ireland. Belgium s situation is more difficult to determine, since that country records only a small number of transfers, all of them under escort. In the opposite camp, the United Kingdom, the Netherlands, Sweden and Denmark are clearly the Member States with the most favourable results as regards the balance of transfers. These observations should be qualified, however, by taking into account (a) the proportion of a Member State s overall demand for asylum that is accounted for by transfers and (b) the influence of factors such as geographical position, which will be discussed later. Nevertheless, despite the variation from one Member State to another, it has to be said that the Dublin Convention does not affect who takes responsibility for examining an asylum application very greatly, since it applies in less than 5% of cases. Unless one assumes that the overwhelming majority of asylum seekers lodge their in the State which the Convention criteria would have designated anyway as the State responsible - which hardly seems to tally with Member States experience of the scale of unlawful movements of third-country nationals within the European Union - the Convention s role as a measure that complements freedom of movement is limited. The causes of this situation are many. The main ones are examined below. 3

II. CRITERIA FOR ALLOCATING RESPONSIBILITY AND THE PRODUCTION OF EVIDENCE There are two aspects to the Dublin Convention: 1) arrangements for determining the State responsible for examining an asylum application and for taking of an applicant; 2) a system whereby the State responsible takes back the asylum seekers for whom it is liable and who are unlawfully on the territory of another State. 1. Establishment of initial responsibility Apart from Articles 4 and 8, which are different in nature, the criteria for determining responsibility (Articles 5, 6 and 7) seek to attribute it to the Member State which originates, or takes the greatest part in, the asylum seeker s entry into or residence in the territories of the contracting States by granting him a residence permit or a visa, by inadequately policing its external borders or by authorising entry without a visa. Only half the Member States are able to provide statistics on the frequency with which each of the criteria is applied. Furthermore, the statistics are not always directly comparable. It is possible, however, to make the following observations on the basis of the cumulative data for the two reference years: The various situations covered by Article 5 (valid or expired residence permit/visa) are both those which are most often referred to in to take and those which give rise to the greatest number, or even the greatest proportion, of positive replies from the Member State approached. Thus Article 5 accounts for 85% of the received by Portugal, 50% of the sent to France in 1999, nearly 50% of the sent to the Netherlands in 1998-99, and 25% of the received by the United Kingdom. It gives rise to one third of the positive replies provided by the United Kingdom and to nearly 60% of the acceptances given by Germany. Article 8 is the second most frequent criterion, which is certainly surprising since it allocates responsibility, for want of other criteria, to the State receiving the application in the first place; one could therefore be faced with a request to take back, as provided for in Article 3(7) and Article 10. However, this seems to indicate an acceleration (which would appear to be confirmed by the fact that use of Article 8 is increasing from year to year) in the secondary movements of asylum seekers who very quickly leave the Member State where they made an asylum application, even before a procedure for determining the State responsible or examining the substance can begin and without formally withdrawing their application as mentioned by the text of the Convention, in order to continue their journey to the Member State where they wish to be. Article 8 accounts for more than 45% of the sent to Belgium, about one third of those sent to the Netherlands, and more than 25% of the received by the United Kingdom. It gives rise to more than one third of the positive replies given by Germany and nearly 30% of the acceptances given by the United Kingdom. In France s case, however, it accounts for less than 1%, which is probably because this Member State interprets most of the sent to it under Article 8 as falling under Articles 3 to 7 or 10. Generally speaking, Article 6 comes in third place as regards the number of to take received by the Member States : 30% for Belgium, 10% for France, nearly 4

10% for Portugal and the Netherlands. However, it accounts for a very small percentage of the acceptances given by the United Kingdom and Germany. Thus, not only is Article 6 cited proportionally less often than one might expect, given the estimates made elsewhere concerning the number of illegal entries across external frontiers, but based on its provisions seem to have a lower success rate than those based on other criteria. Article 4 is used very little: in two years it has concerned one case out of 286 in Portugal, 16 out of 961 in Belgium, 20 out of 295 in the United Kingdom and 64 out of 1 464 in the Netherlands. This seems to indicate that, if anything, the family members of refugees residing in the Member States go through the regular channels for bringing families together and that, where this is not the case, there are few who have to make out an application for asylum en route in a Member State other than that where the refugee resides. Lawful entry as referred to in Article 7 seems to be the criterion cited least often: 37 cases in two years in the Netherlands, 18 in Belgium, 11 in the United Kingdom, and 9 in Portugal. It accounts for 7% of the acceptances given by the United Kingdom and barely more than 1% of the positive replies from Germany. The relative frequency with which the various responsibility criteria are applied is determined partly by the numerical importance of each of the situations referred to and partly by how easy it is to establish that an asylum seeker is in the situation described. However, it would be possible to verify whether, for each Member State, there is a proportional link between its action on asylum and the movement of persons and the frequency of use of each of the Convention criteria in that respect only by comparing a large number of associated statistical data, which would be outside the scope of this survey. It is possible, however, to take account of the impact of the production of evidence: As regards Article 4, it is sufficient for the Member State presumed to be responsible to check in its records whether a particular person does in fact reside as a refugee on its territory. Provided the information supplied by the asylum seeker is correct and the family link is established, such a check does not really pose a problem; thus, if anything, the reasons why this criterion is not used very much should be sought in the remarks set out above. Nevertheless, strict requirements as regards evidence of the family link or of the matrimonial link and refusals to take account of marriages subsequent to the departure of the persons concerned from their country of origin may have helped to restrict the scope of this provision. As far as Article 5 is concerned, the replies from the Member States show that this criterion is relatively easy to apply. There can be hardly any dispute where a tangible object (residence permit, visa) is available, although some instances of fraud (assuming somebody else s identity, forging the document) have been interpreted differently. Where no such object is available, the criterion may, most often, be applied only if the asylum seeker himself provides the information which can guide the search: the Member State presumed to be responsible may then be questioned and carry out a check in its records. This is especially true as regards visas; however, it seems that, given the differences in organisation, Member States are not equally 5

effective when it comes to checking the issuance of a visa by one of their representations abroad. It is with regard to the application of Article 6 that the difficulty of providing evidence that the border has been crossed illegally is keenest, since, generally speaking and almost by definition, a clandestine operation leaves no official trace. The search for the State responsible can be guided only by the declarations of the asylum seeker or the presence of indicative evidence. In practice, an asylum seeker s declarations are regarded as evidence of low probative value: they are taken into account only if they are accurate and detailed and contain information which can be checked. Similarly, indicative evidence which supports the presumption that the asylum seeker has traversed the territory of a Member State is assessed differently according to whether it contains a name (invoice, plane ticket) or not (train ticket, currency, etc). After some initial hesitancy, a sort of common understanding seems to have been established regarding the standard of evidence needed to invoke Article 6 vis-à-vis another Member State: thus, the replies from several Member States show that the competent authority is often forced to forgo informing the State presumed to be responsible where, despite the conviction that that State is the one through which the asylum seeker has entered, there is insufficient evidence. However, this self-restraint, which leads to another State being informed only if there is a definite basis for doing so, is not enough to ensure the success of a request to take : the available statistics show that the rate of acceptance is still lower than for other criteria (the United Kingdom, for example, accepted only three out of the 16 submitted to it under Article 6, whereas the data supplied by Germany show that, for 1998-99, there were 226 acceptances as against 465 rejections). As regards Article 7, the problem of evidence arises in virtually the same terms as for Article 6. The best evidence, i.e. the stamp placed in the passport on entry, seems to be rarely available, since it depends both on a passport being submitted by the asylum seeker and on the systematic placing of an ink stamp by the authorities responsible for controlling the authorised points of entry. The use of other types of probative and initiative evidence is mutatis mutandis the same as for Article 6. In principle, all for asylum are registered. Thus, as soon as the declarations of the asylum seeker or any tangible item of evidence make it possible to steer inquiries in a particular direction, it is easy to contact the Member State presumed to be responsible, so that it checks in its records whether there is such an application with a view to applying the Article 8 criterion, provided the case is not one where the applicant should be taken back under Article 3(7) and Article 10. Bilateral exchanges of fingerprints are used successfully to overcome the difficulties that may result from changes in the spelling of names (a problem frequently associated with the transliteration of patronymics written in another alphabet). This rapid review of the constraints attaching to the production of evidence shows that, in most cases, the asylum seeker himself is fundamental to the ability of a Member State to determine that another Member State is responsible. He may, through his silence and/or the suppression of evidence, deprive the Member State to which he is applying of any means of action under the Dublin Convention. The Member States are divided about whether the entry into force of the arrangements for determining the State responsible has brought about an increase in the proportion of asylum seekers without documents: some believe that this phenomenon already existed, others believe that the trend is rising but 6

cannot quantify it. The data available do not enable us to determine the extent to which the Member States administrative organisation or the welcome they provide for asylum seekers affects their application of the Convention, in particular by facilitating, to varying degrees, complete, early access to the evidence which the asylum seeker may carry on him or in his luggage. 2. Taking back Taking back concerns either returning to the Member State where he lodged his application an asylum seeker who has withdrawn his application and has gone to another Member State while the State responsible is being determined (Articles 3 to 7), or returning to the State responsible an asylum seeker who has gone to another Member State when his application is being examined (Article 10(1)(c)), has been withdrawn (Article 10(1)(d)) or has been rejected (Article 10(1)(e)). It can be regarded as a readmission mechanism. Although taking back is not recorded separately by all Member States, the data available shows that it is a major part of the activity associated with the application of the Convention. It accounts for more than 50% of cases in some Member States, viz. 224 of the 308 sent by Finland to other States in 1998-99, and 90 of the 176 received by that State; 89 of the 195 made by Luxembourg in 1999, and 16 of the 22 received by the Grand Duchy. In Germany, taking back accounts for more than 25% of the acceptances given to from other Member States (4 592 out of 16 915 in 1998-99). There was, however, a sharp drop from the first year to the second (3 409 out of 9 263 in 1998, 1 183 out of 7 652 in 1999). This was offset by an opposite trend as regards based on Article 8 (which went up from 1 303 to 2 990). Thus, the of cases in which the responsibility of the Federal Republic results from the fact that the asylum seeker started a procedure in that country remains relatively stable. In the United Kingdom, cases covered by Article 10 account for about 20% of the total, both as regards incoming cases (i.e. sent to the UK) and outgoing cases (originating in the UK). Austria is a special case, since incoming cases (1 107 out of 3 523) are much more numerous than outgoing cases (69 out of 1 536). Lastly, in Portugal, the application of Article 10 covers only one case a year. The replies from the Member States show that it is generally easier to provide evidence of the need to take an applicant back than to demonstrate the initial responsibility of another Member State. The recording of asylum in reliable, centralised files and the possibility of exchanging fingerprints on a bilateral basis largely explain these results. The introduction of EURODAC, which renders the identification of multiple virtually infallible and facilitates the detection of submitted earlier by an alien who is unlawfully present in another Member State, should considerably increase the number of applicants taken back. Although this is not objectively confirmed by statistics, the general impression obtained from the Member States replies is that taking back results more frequently in an effective transfer of the asylum seeker than does taking. Applicants, it seems, are more ready to go back to the Member State where their application is being examined, or has been rejected after a procedure whose duration has allowed them to get used to the Member State responsible, than to go to the State responsible and submit their initial 7

application there when it is not their chosen destination. Another explanation is that, in several Member States, persons who are the subject of a request to take back are kept in detention and often transferred under escort. 3. Requests for information under Article 15 As indicative evidence (whether tangible items or declarations by the applicant) is generally insufficient, confirmation of a presumption may be sought by questioning the Member State presumed to be responsible pursuant to Article 15. The State questioned is, in principle, obliged to carry out the necessary checks as far as possible, in particular in its records, and to provide an answer in a spirit of reasonable cooperation. An affirmative answer to a question concerning the issuance of a residence permit or visa will in most cases be sufficient evidence of the responsibility of the State concerned. Only eight Member States were able to provide complete statistics on the number of for information issued and received by them, and only three of these provided precise information on the success rate for those. Four other Member States provided partial data or estimates (see Table IV). The following observations can be made on the basis of the data available: The number of for information increases considerably in the second year in all Member States except Spain, which issued fewer in 1999 than in 1998 (7 instead of 40). In some cases the increase is spectacular: Austria, 123 issued in 1998, 2 083 in 1999; Ireland, 1 468 then 8 411; Belgium and the United Kingdom, 16 000 then 25 000, and so on. A similar trend is perforce observed as regards received. The success rate for for information declines in the second year: for Austria it falls from 35.80% in 1998 to 19.30% in 1999, for Ireland from 29.50% to 8% and for the United Kingdom from 11.40% to 8.20%. The replies from the Member States show that the processing of for information constitutes a considerable administrative burden, in particular for the State which receives them. The time it takes to reply, which is sometimes very long, has an impact on the average duration of the procedure for determining the State responsible and may in some cases adversely affect the smooth operation of the Convention. The ability of the Member States to handle the for information sent to them depends partly on administrative organisation (arrangements for access to various databases), but even more on the resources which they are able to devote to them. The proliferation of for information raises questions about the existence of non-targeted, i.e. universal issued more or less at random, which clog the system and impair overall efficiency. 4. Exemption clauses The obligation on the State responsible to take or take back ceases under certain conditions laid down by the Convention, i.e. if the request to take is submitted more than six months after the asylum application was lodged; if the asylum seeker has resided for more than six months in a Member State after his unlawful entry across the external border of another Member State; if the applicant has left the territories of the Member States for more than three months or has been effectively removed. 8

Experience shows that, most often, such exemptions are difficult, if not impossible, to prove. In particular, it is virtually impossible, unless the applicant provides evidence himself, to verify or invalidate the declarations of an asylum seeker concerning his exit from the territories of the Member States and the duration of his absence, if the latter is not the result of a removal carried out by the Member State. III. TIME LIMITS One of the objectives of the Dublin Convention is to avoid leaving asylum seekers uncertain about the outcome of their application for too long. The achievement of this objective is compromised, if the time limits placed on the procedures for determining the State responsible are excessive. 1. Time limits laid down by the Convention and by decision No 1/97 of the Article 18 Committee of 9 September 1997 Article 11(1) of the Convention grants the Member State with which an asylum application has been lodged a maximum period of six months from the date of application in which to request another Member State that it should take. A small minority of Member States considers that this time limit is too long to be compatible with the objective of speed established in the Convention in the interest of asylum seekers. Several Member States, on the other hand, think that the time limit is appropriate, or even too short, given that the evidence of another State s responsibility may emerge only late in the procedure and that the time limits for replying to for information sent to other Member States under Article 15 may frequently mean that the six-month time limit is exceeded. Article 11(4) stipulates that the State to which a request to take has been made should pronounce judgment within three months, failure to do which implies acceptance. Most of the Member States consider this time limit to be excessive and want the optimum time limit of one month mentioned in Article 4(1) of decision No 1/97 of the Article 18 Committee, which seems realistic and is observed in the majority of cases, to become a binding maximum. The replies from the Member States show, however, some cases where the three-month time limit is exceeded and emphasise that, in such a scenario, it is virtually impossible to force the defaulting State to assume the consequences of its implicit acceptance. However, the time limit of one month provided for by Article 11(5) and Article 13(1)(b) for the transfer of the asylum seeker is generally perceived as too short for notifying the decision to the requester, then organising and carrying out a transfer in all cases. It should be noted, though, that the time limit has only indicative value in practice, since the Member States have agreed in Article 21 of the above-mentioned decision of the Article 18 Committee that the absence of an effective transfer does not alter the obligations of the State responsible, which is still bound at all times to take of/take back the applicant whose transfer has not taken place (in consultation with the requesting State and notwithstanding the provisions of Article 10(2), (3) and (4) of the Convention). 9

Assuming the various stages of the procedure are accomplished within the maximum time limits laid down by the Convention, the time expiring between registration of an asylum application in a Member State and transfer to the State responsible could be as much as ten months. In practice, the time required for the competent authorities to act is generally less, in particular because most transfer are made well before the expiry of the six-month time limit. As regards that an applicant be taken back, Article 13 lays down a time limit of eight days for answering the request submitted by another Member State and, as with to take, a time limit of one month for the transfer. Many Member States consider that this time limit is too short, in particular where fingerprints have to be checked, but consider that the introduction of EURODAC should resolve this problem. Adopting a longer time limit would create problems for those Member States where the law does not allow the detention of aliens in an illegal situation to be extended. 2. Other sources of time limits As indicated above, the time limit on replying to a request for information under Article 15, which is not regulated by the Convention, is frequently regarded as a reason why procedures are prolonged, or even why the six-month time limit in Article 11(1) is exceeded. Various suggestions have been made in this respect: the time limit provided for by Article 11(1) should be extended, or its operation suspended pending a reply to the request for information. Such solutions would only help to increase the duration of the procedures without guaranteeing improved efficiency and seem hardly compatible with the objective of speed stated in the Convention; a maximum time limit should be set, possibly coupled with a penalty, for replying to a request for information. The risk inherent in this situation - greater proliferation of for information and hence obstruction of the competent authorities ability to handle them, to the detriment of the processing of to take /take back - could only be avoided, unless the resources of the departments concerned are increased, by provisions narrowly defining those cases where a request for information may be sent to another Member State. Moreover, the asylum seeker himself may, through his actions, contribute to prolonging the process: by lodging a suspensive appeal against the decision affecting him taken under the Convention (in those Member States where an appeal has, or may have, such an effect); by evading the enforcement of the transfer decision. Lastly, circumstances such as illness, pregnancy and maternity may also increase the time taken. 3. Consequences of time limits There is general agreement among the Member States that the longer the procedure, the less chance there is that the transfer will actually take place. 10

The existing data on average time limits is too fragmentary to allow calculation of the average duration of the procedure for determining the State responsible; the data compiled quarterly relate only to the average time elapsing between a request to take /take back and the reply from the Member State, and do not make it possible by any means to assess the period between registration of an asylum application and the sending of a request to take, nor the period between acceptance of responsibility and the transfer of the applicant to the State responsible. While it is possible to envisage adjustments to the time limits laid down in the Convention, these can have only a marginal effect, given the importance of causes outside the Convention itself. It is clear that the excessive extension of certain procedures is due principally: either to physical constraints, on which legal provisions have little or no effect: difficulty in collecting and assessing evidence, ability of departments to process for information and to take /take back, unforeseeable circumstances such as the asylum seeker s state of health; or to constraints associated with the national law of the Member States, in particular the existence of suspensive appeals and the duration of court proceedings. It is clear, moreover, that in certain Member States the time limits determined by national law for inter alia the duration of the detention or the maximum period for taking a decision about the admissibility of the asylum application are shorter than those stipulated by the Convention or the reply times found in practice. The result is that the Member State concerned is sometimes unable to continue the procedure begun in accordance with the Convention, either because the applicant has to be released before the transfer is carried out or because, for lack of a reply from the State requested, the requesting State is itself forced to assume responsibility for examining the application. In any event, for the asylum seeker, the procedure for determining the State responsible, whatever its duration and outcome, is additional to the procedure for deciding eligibility for refugee status. For the Member State conducting it, a determination procedure does not replace the eligibility procedure. It renders it unnecessary, only if the State requested agrees to acknowledge responsibility; if that State does not, the first Member State will also have to carry out an eligibility procedure, which requires extra time. 4. Role of bilateral agreements Several Member States have concluded bilateral agreements making it possible to simplify and accelerate application of the Convention in certain situations. The agreements concluded by Germany with Denmark, Sweden and Austria are intended where possible to settle within a very short time frame cases occurring during checks at entry points and in border areas. In derogation from normal procedure, cases where the responsibility of the other party for taking back/taking can be established immediately are processed within a time limit agreed directly between the authorities responsible for controlling the borders, or between the latter and the central authority competent for determining the State responsible. The statistics of Germany s Federal Office for the Recognition of Foreign Refugees clearly show the numerical importance and efficiency of these procedures in 1999: 11

vis-à-vis Denmark: Germany presented 19 under the agreement out of a total of 81; acceptance was granted in 18 cases, all of which resulted in an effective transfer. In the opposite direction, out of the 2 964 which it sent to Germany, Denmark submitted 2 109 under the agreement and received 475 acceptances, which resulted in 474 transfers ; vis-à-vis Austria, the figures were 989 out of 1 603, 696 acceptances and 620 effective transfers. IV. LEGAL QUESTIONS In applying the Convention, the Member States have been faced with unforeseen situations likely to complicate, slow down, limit or prevent the operation of the arrangements for determining the State responsible and for taking /taking back. 1. Vagueness of, and gaps in, the Convention Some of the difficulties encountered are due to the Convention itself. The chief ones are examined below: a) Scope Given the definitions in Article 1(1)(b) and (c), the Convention provisions relating to the determination of the State responsible apply only to third-country nationals who are seeking recognition of their refugee status under the Geneva Convention of 28 July 1951 and in respect of whose no final decision has been taken. The provisions relating to taking back apply both to such applicants and also to persons who can no longer be described as asylum seekers following rejection of their application (Article 10(1)(e). The result is, firstly, that the Convention does not apply to aliens who apply for protection on grounds other than the Geneva Convention. Some aliens have sought to evade the application of the Convention and to stay in the non-responsible State where they lodged their claim, either by withdrawing their request for refugee status while maintaining or expressing a request for protection on different grounds, or by immediately drawing up their request for protection on grounds other than the Geneva Convention. Member States were asked, first, whether an asylum seeker who changes the nature of his application for protection can still be covered by the Convention and, second, whether it is advisable and feasible to extend the scope of the instrument (or its successor) to include for alternative, complementary or secondary forms of protection. The answer to the first question was negative. On the second point, it seemed that the disparity of the alternative forms of protection to be found in the Member States made it difficult to conceive of this exercise as long as a degree of harmonisation had not been achieved (this was also the conclusion of the Commission s services in working paper SEC (2000) 522). 12

It should be noted that this escape route to another form of protection is possible only if the national law of the Member State allows people to apply for such protection separately from that provided for by the Geneva Convention. In Member States which have a one-stop shop system, i.e. where the application relates simultaneously to all forms of protection provided for in law and where it is up to the competent authorities to determine which type of protection the applicant is eligible for, this escape route is generally not possible. In any event, the scale of this phenomenon is still limited. The second result is that the provisions of the Convention are regarded as inapplicable to an alien who makes an application for asylum in a Member State when the status of refugee has already been granted him in another Member State. This situation, which occurs in practice but whose frequency is impossible to measure (see the provisions of the EURODAC Regulation on the blocking of the fingerprints of recognised refugees, for (provisionally) statistical purposes), could possibly be settled by other means. This is, however, a matter of interpretation, about which the Member States are not unanimous. b) Definitions Application for asylum : despite the comments set out in Article 2 of decision No 1/97 of the Article 18 Committee, there is still a vagueness and lack of uniformity, associated with differences of procedure in the Member States, as to what formally constitutes an application for asylum within the meaning of the Convention; this may have consequences for calculating certain time limits and for the application of the provisions on taking back. Examination of an asylum application : also because of differences in procedures, it is not always easy to determine how far action by the competent authorities may be regarded as limited to determining the State responsible, or whether it amounts to an examination of the substance of an application, such as to involve the responsibility of the Member State and, where appropriate, preclude, pursuant to Article 3(4), the possible responsibility of another State. Residence permit : there is no uniform assessment of the impact on the Convention of the nature of certain types of residence/tolerance equivalent to secondary protection. The concept of withdrawal of the application used in Article 3(7) and Article 10(1)(d) should be clarified; in practice there are more cases of an application being abandoned than formally withdrawn. The concepts of irregular in Article 10(1)(c) and illegal in Article 10(1)(e) are not precise enough to guarantee uniform interpretation of the situations to which they refer. c) Wording of the provisions Several Member States say they are perplexed by the confusion caused by Articles 3, 8 and 10; Given the uncertainty surrounding the expression examination of an application for asylum, it may be difficult to invoke with regard to another Member State the 13

responsibility that arises from use of the sovereignty clause in Article 3(4) and to situate this in the hierarchy of criteria. The provisions of Article 7 take no account of the realities resulting from application of the Agreement implementing the Schengen Agreement; the contracting States to this Agreement have had to introduce a special arrangement adopted by decision of the Executive Committee. This situation is not really compatible with the place of the Schengen acquis in the new edifice created by the Treaty of Amsterdam. There is clearly an overlap in practice between Article 8 and the situations provided for in Articles 3(7) and 10(1), highlighted by the fact that Article 3(7) does not figure in the Member States statistics on use of the criteria. The effects of these provisions are not the same: taking back under Article 3(7) does not prejudice responsibility, whereas taking /taking back under Article 8 means that the State which agrees to do so admits responsibility for examining the application. d) Gaps in the Convention Unity of families : apart from the question of scope, examined above, the entirely discretionary nature of Article 9 and the lack of explanation concerning the humanitarian reasons of a cultural or family nature are regarded by many Member States as a shortcoming which impedes the smooth operation of the Convention. The examples given, however, concern cases of family unity and do not come under Article 4. It may be concluded that the lack of clear, binding autonomous provisions making it possible to combine respect for unity of the family with the principles underlying the criteria for allocating responsibility is indeed a gap in the Convention, to which decision No 1/2000 of the Article 18 Committee formalising the principles on which the Member States were already basing their activity, provides only a partial solution. Interpretation : many of the questions raised in the above paragraphs (but also in section II) are the result of differences of interpretation. The Article 18 mechanism has proved clumsy and the Member States have hesitated to use it for questions which do not concern a considerable number of cases. The lack of a mechanism for interpreting questions/settling disputes of a legal nature is regarded by many as a shortcoming. 2. Action by national courts Since they may give rise to complaints, decisions to transfer an asylum seeker from one Member State to the State responsible may be appealed under the conditions laid down in national law. In States where such an appeal is not suspensive and there is, hence, little advantage to be gained from taking action, few appeals are lodged and case law on the application of the Convention is not extensive, or is even non-existent. However, in those Member States where appeal has, or may have, suspensive effect, litigation is more common and has resulted in several legal decisions which are likely to impose further constraints on the manner in which those States may apply the Convention. The information supplied by the Member States which have data on this subject show that the number of cases referred to a review body increases in the second year, both in absolute terms and as a proportion: in Austria, the number of appeals against transfer decisions went up from 6 in 1998 to 358 in 1999; in Germany, from 83 to 200; in Denmark, from 132 to 208; and in the United Kingdom, from 376 to 484. In the Netherlands, 80% of decisions are referred to the courts with a view to obtaining a stay of 14

performance which will make it possible for the person to remain in the country pending the result of an appeal. Average duration varies considerably from one Member State to the next: the Austrian law lays down a period of 10 to 20 days, while in Sweden the average duration is three weeks; in Germany it is 130 days, and in Ireland the time allowed was reduced from 16 months in 1998 to eight months in 1999; in the United Kingdom the average period is 18 months, with considerable variation depending on whether the person is detained (three months on average) or not (approximately two years). Most of the case law focuses on the application of Article 3(4), firstly as regards its relation to certain provisions of national law (constitution, reasons for administrative decisions, information given to the applicant, the authority s duty of diligence, etc.), and secondly as regards international law - principally how it relates to Article 8 of the European Convention on the Protection of Human Rights and Fundamental Freedoms (ECHR), Article 3 of the said Convention and Articles 1 and 33 of the Geneva Convention relating to the Status of Refugees. While it is not possible in this report to go into the details of the various court decisions, iii it may be noted that: generally speaking, the competent courts have considered that the provisions of the Convention do not create individual rights which appellants may invoke to their advantage, but in a good many cases they have held that the Member State was nevertheless obliged to take account of the consequences of an inadmissibility decision based on the Convention - and hence of the non-applicability of the discretionary power granted by Article 3(4) to examine the question of responsibility for an application where that State is not responsible itself - on the situation of the applicant as regards respect for family life (Article 8 ECHR) or as regards the principle of non-refoulement (Article 33 of the Geneva Convention) and the prohibition of torture and inhuman and degrading treatment and punishment (Article 3 ECHR). The effect of these decisions is to force the Member States concerned to carry out, in addition to determining the State responsible, an assessment in each case of factors quite outside the Dublin Convention criteria and, where appropriate, to state explicitly the reasons why they are not resorting to Article 3(4). The individual assessment is likely to make the decision-making process of the competent authorities considerably more complicated. Particularly is this so where the assessment concerns how one Member State interprets or applies the Geneva Convention or grants the benefit of other forms of protection, compared with the interpretation/prevailing practice in the requesting Member State. In this respect, it is clear that the position of principle adopted by the courts in the United Kingdom as regards interpretation of the Geneva Convention in cases of persecution by non-governmental agents in France and Germany is likely to force the UK authorities to consider, in a number of cases, the very grounds of an asylum application in a way which is not all that different to an examination of the application s substance. Article 15B(1) of the Dutch Law on Immigration, known as the tenzij (unless) clause is likely to have the same consequences. 15